Unfamiliar ground is where I do my best work.
I work on problems that are messy, consequential and don’t belong to one discipline: entering an unfamiliar market, a project that has stalled, a decision buried in documents. I find out what is actually happening, then turn it into something you can act on.
I read situations the way a historian reads a source.
History trained me to ask where a document came from, who wrote it, for whom, and what it was meant to achieve. To tell what people say from what they mean, and from what is making them say it. Patterns, positions, culture, power. I apply the same questions to a regulation, a market, a partnership or a project.
- What does the rule say?The official version: the law, the policy, the brief, the pitch.
- Who actually has to act?Rarely the people who wrote it.
- What are their incentives?Money, reputation, politics, survival.
- Who absorbs the cost?Usually whoever has least power to refuse.
- What happens next?A dated call I can be held to.
Much of what matters isn’t in the file. So I create information as well as analyse it: I call people, ask the second question, walk the site, and compare accounts. I keep evidence, inference and assumption separate, and I say how confident I am. AI lets me do far more with what I find. It doesn’t do the finding.
It is a question I have been asking for fifteen years: of the international response to Somali piracy in 2012, of who pays for supply-chain compliance in 2018, and now of the signals in The Ledger.
Egypt, 2013. The engineering was sound. The customer was wrong.
The question looked technical: what size of solar pump suits a desert farm? Aquifer depth, crop water, sunshine. That part could be answered.
But farmers bought diesel on subsidy coupons at LE 1.10 a litre. At that price solar took more than twenty-five years to pay back for the farmer, while the state quietly covered the difference. The real case was to the government: solar as a way to cut the subsidy bill.
Today the subsidies are largely gone, and the same system pays back in about six years. The physics is the same; the incentives changed.
Most stuck projects look like this: a problem defined as technical that is really about who pays, who decides, or who has to act.